Modulating analytical characteristics of thermovinified Carignan musts and the volatile composition of the resulting wines through the heating temperature Olivier Geffroy, Ricardo Lopez, Carole Feilhes, Frédéric Violleau, Didier Kleiber, Jean-Luc Favarel, Vicente Ferreira

To cite this version:

Olivier Geffroy, Ricardo Lopez, Carole Feilhes, Frédéric Violleau, Didier Kleiber, et al.. Modulat- ing analytical characteristics of thermovinified Carignan musts and the volatile composition of the resulting wines through the heating temperature. Food Chemistry, Elsevier, 2018, 257, pp.7-14. ￿10.1016/j.foodchem.2018.02.153￿. ￿hal-02373447￿

HAL Id: hal-02373447 https://hal.archives-ouvertes.fr/hal-02373447 Submitted on 21 Nov 2019

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés.

OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible

This is an author’s version published in: http://oatao.univ-toulouse.fr/25066

Official URL: https://doi.org/10.1016/j.foodchem.2018.02.153

To cite this version:

Geffroy, Olivier and Lopez, Ricardo and Feilhes, Carole and Violleau, Frédéric and Kleiber, Didier and Favarel, Jean-Luc and Ferreira, Vicente Modulating analytical characteristics of thermovinified Carignan musts and the volatile composition of the resulting wines through the heating temperature. (2018) Food Chemistry, 257. 7-14. ISSN 0308-8146

Any correspondence concerning this service should be sent to the repository administrator: [email protected] Modulating analytical characteristics of thermovinified Carignan musts and the volatile composition of the resulting wines through the heating temperature

3 b C a h h Olivier Geffroy · •*, Ricardo Lopez , Carole Feilhes , Frédéric Violleau , Didier Kleiber , d c Jean-Luc Favarel , Vicente Ferreira

a lnstllutFronçais de la Vl&ne et dtVin Pôle Sud-Ouest, V'lnnopôle, BP22, P-81310Lisle SurTom, France b lJni..,,-sitide Toulouse, &oie tflnginlnusde Putpa1!, LaboraJOtre d'Agro-PhyslologJe, 7S voiedu TOEC, BPS7611,P-31076 Toulouse Cedex3, Pronœ C LaboraJOryPlavor for Analysis and Enology,lnstltuto Agroa1imentarlo de ArQ&!Sn (1A2),D,partment of AnalytkalCheiristry, of Paculty Sdences, Uni"'1'SidadZ

ARTICLE INFO ABSTRACT

Keywords: Theimpact of two temperaturelevels (50 •c and 75 °C) and heatinglimes (30 min and 3 h) on thecomposition of Thennovinlflcation thermovinified musts and winesfrom Carignan was investigated at the laboratory scale in 2014 and 2015. The Pre-fennentation heattreatment heating temperature had a significant impact on the extraction of amino acids and a probable thermal de­ Heatlngtemperature gradation of anthocyanins was noted at 75 •c. ln 2014, musts from grapesthat underwent a heat treatment at Heatlngtlme 50 •c for 3 h had a similar level ofphenolic compoundsa s those treateda t 75 •c for30 min. This indicates that Aroma composition the reduction ofthe heating temperature in some vintagescan becompensated for through an extension of the ln Chemkalccmpounds studled tfusarticle: heatingperiod. Severa! grape-derived molecules were impa ctedby the rise in temperature and winesmade from f,-Damascenone(PubChem 01>. 5374527) grapes treated at 50 •c in most casescontained largerconœntr ations of geraniol, j3-citronellol, j3-damascenone f,-Cltronellol (PubChem OD:8842) a-Terpineol (PubChemOD: 17100) and 3-mercaptohexanol. Geranlol (PubChem OD:637566) 4-Mercapto+methyl-2-pentanone (PubChem OD: 88290) 3-Mercaptohexanol(PubChem OD: 521348) Guaiac:ol (PubChemOD: 460) 2-Methyl-3-furanthlol (PubChem OD:34286) y-Nonalactcne(PubChem 01>. 7710).

1. Introduction persona! communication, July 18, 2017). In comparison with control macerated wines, thermovinification Pre fermentation heattreatment of grapesor thermovinification is a wines fermented at a lower temperature usually have higher levels of winemaking technology first industrially developed in the seventies ethanol (Geffroyet al., 2015). Sorne changes in the acid base balance of (Marteau & Olivieri,1970 ). lt consistsof heating grapes between70 and the wines by higher tartaric acid and potassium extractions from the 75 •c for a length of rune varying from 30 min to 24 h. The term pericarp tissueof the berries were also highlighted. The typicalsensory "thermovinification" is sometiines used to describe the processin which profile of thermovinified wines is due to a large extraction under the heating is limited to a brief period ( < lh). After pressing at a high effects of heat of hydrophilic anthocyanin pigments and grape poly temperature and clarification, fermentation is usually undertaken in saccharides responsible for roundness in wine (Doco, Williams, & liquid phase at a lower temperature than usual red ferments, typically Cheynier, 2007; Girard, Kopp, Reynolds, & Cliff, 1997). In aqueous between 18 •c and 25 •c. This technique is becoming increasingly phase, the extraction of tannins is more moderate and the technique popular for the production of colored, fruit driven red wines with soft usually leads to wines with a lower overallphenolic content than those tannins. The volume of wine elaborated in France in 2017 through made using traditionalmaceration techniques (Auw, Blanco, O'Keefe, & thermovinification was esti inated at 750 million liters (J.L Favarel, Sims, 1996).

• Correspondlngauthor at: Université de Toulouse,&:oie d1ngénieurs de Purpan, Laboratoired'Agro-Physlolog!e, 75 vole du TOEC, BP57611, F-31076 ToulouseCedex 3, France. E-mailadtt-ess: [email protected](O. Geffroy).

https://doi.org/10.1016/j.foodchem.2018.02.153 From an olfactive point of view, thermovinification is known to with 2.40 m × 1.40 m vine spacing and a moderate production level produce wines with a standardized sensory profile often described as (6 8 t/ha). 10 kg of grapes were hand harvested on 6 October in 2014 “banana yogurt” by winemakers. Past research into the volatile com and on 25 September in 2015 in 1 case of 20 kg. The grapes were then position of thermovinified wines showed that maceration heat treat destemmed manually, mixed gently and divided into eight homogenous ment allowed the elimination through volatilization of a large amount lots of 1000 g. of 3 isobutyl 2 methoxypyrazine (Roujou de Boubée, 2000) and was not favorable to the production of wines with high concentrations of 2.2. Pre fermentation heat treatments hydrophobic rotundone (Geffroy, Siebert, Silvano, & Herderich, 2016). Another work reported the likely thermal degradation of several grape In 2014 and 2015, each pre fermentation heat treatment was re derived aroma compounds or their precursors (i.e. some varietal thiols, plicated twice. 1000 g of berry samples were crushed and poured into a monoterpenols, norisoprenoids, phenols) when grapes were heated 1 L Erlenmeyer flask (with a perforated lid to evacuate carbon dioxide during 3 h at 70 °C (Geffroy et al., 2015). The fermentation conditions during fermentation), and sulfur dioxide (40 mg/L) was added using a of thermovinified wines particularly enhanced esters, acetates and fatty 10% bisulfite liquid solution. The heating was carried out at two tem acid formation (Cottereau & Desseigne, 2007; Fischer, Strasser, & perature levels (50 °C and 75 °C) using a water bath system. The rise in Gutzler, 2000; Girard et al., 1997). Consistent with previous observa temperature of the grapes from room temperature up to the desired tions made in the seventies (Poux, 1974), Geffroy et al. (2015) recently temperature was fast, taking exactly 40 min. For each target tempera showed that heating at 70 °C for 2 h followed by pressing at a high ture, the heating was maintained for 30 min for 2 out of 4 samples and temperature induced a substantial increase in the concentration of for 3 h for the remaining flasks. The grapes were then pressed at a high amino acids in the must (from + 101% to 200%). The fermentation in temperature under controlled conditions (200 kPa for 2 min) using a liquid phase and at low temperature of high Yeast Assimilable Nitrogen laboratory press (Paul Arauner GmbH, Kitzingen, Germany). The (YAN) clarified musts enhanced the production of fermentative aroma weight of must at pressing was measured and the extraction rate (%) compounds by the yeast (Moreno, Medina, & Garcia, 1988). expressed as the weight of must (g) obtained from the pressing of 100 g The modulation of the sensory profile of thermovinified wines to of berries was calculated. The musts were centrifuged (14,000×g for wards a fruiter varietal character is an issue frequently raised by 6 min) and 200 mL were sampled to perform classical enological ana winemakers. Grape derived aroma compounds imparting this character lysis and determination of polyfunctional thiols precursors. To avoid to the wine include monoterpenes, norisoprenoids, aliphatics, phenyl any bias due to distinct levels of clarification between the studied propanoids, methoxypyrazines, and volatile sulfur compounds treatments, turbidity was controlled using a 2100AN turbidimeter (Robinson et al., 2014). (Hachlange, Düsseldorf, Germany). After centrifugation, differences in In an attempt to produce thermovinified wines with a fruiter var turbidity between the samples were weak; the average value was 87 ietal character, two levels of must clarification (150 and 800 nephelo NTU ± 13. The musts were then inoculated with 200 mg/L of rehy metric turbidity units or NTU) and fermentation temperature (18 °C and drated active dried Saccharomyces cerevisiae yeast (Anchor NT116®,La 25 °C) were previously investigated (Geffroy et al. (2014)). The results Littorale, Servian, France). To promote the production of wine with a were inconclusive as both factors had an overall weak impact on the varietal and complex sensory profile, the musts were fermented at 25 °C aroma composition and sensory profile of the wines. However, wines for 12 days. The kinetics of fermentation was monitored daily by fermented at 25 °C were judged slightly less amylic and more complex. manual weighing of the flasks. After that period, the wines were cen Minor differences were observed between wines made from distinct trifuged (14,000×g for 6 min) and received a sulfite addition of 80 mg/ clarification levels, and these were limited to mouthfeel and taste L. After bottling into 200 mL bottles, the samples were stored at 4 °C perception. until the aroma composition analysis. In relation with amino acid extraction, thermal degradation and volatilization of aroma compounds, the heating temperatures could be 2.3. Must analysis adjusted to modulate the volatile composition of thermovinified wines. Thermovinification was originally used on botrytized grapes to destroy 2.3.1. Conventional enological analysis laccase whose activity increases with temperature up to its denatura Conventional enological parameters were determined after one day. tion point of 60 °C. This is the main reason why the heating of rotten The sugar concentration (°Brix) was determined with a digital hand grapes must be done very quickly at a temperature above 70 °C held Pocket refractometer PAL (Atago, Japan) and the pH with a (Ribéreau Gayon, Dubourdieu, Donèche, & Lonvaud, 2005). Nowadays Titromatic pHmeter (Hachlange, Düsseldorf, Germany). The titratable thermovinification is mainly employed on grapes of perfect sanitary acidity was measured following the OIV method (2009). A Konelab status without Botrytis cinerea and new ranges of temperature, espe Arena 20 sequential analyzer (Thermo Electron Corporation, Waltham, cially below 60 °C, deserve to be investigated. As a decrease in the USA) associated with enzyme kits provided by several suppliers was heating temperature is likely to impact the level of phenolic compounds used to determine amino acids, ammonium (Megazyme, Ireland) and in wine, the heating time would need to be adapted. malic acid (Thermo Fisher Scientific, Waltham, USA). Potassium de The purpose of the present work is to study the impact of the termination was done by flame photometry (Bio Arrow, France) fol temperature and the heating time on the analytical and volatile com lowing the OIV method (2009) and tartaric acid determination by position of thermovinified musts and wines. In 2014 and 2015, two colorimetric titration (Hill & Caputi, 2009). Anthocyanins and the Total temperature levels (50 °C and 75 °C) and heating times (30 min and 3 h) Phenolic Index (TPI) were quantified following the techniques de were investigated in duplicate at the laboratory scale on Carignan scribed by Ribéreau Gayon and Stonestreet (1965) and Ribéreau Gayon grapes sourced in Spain. (1970), respectively, using an Evolution 100 spectrophotometer (Thermo Electron Corporation, Waltham, USA). Absorbance was mea 2. Material and methods sured at 420, 520 and 620 nm and Color Intensity was calculated by summing the three color components (A420 yellow, A520 red, and 2.1. Grapes and vineyard location A620 blue). All determinations were carried out in duplicate.

The experiment was carried out with Vitis vinifera L. cv. Carignan 2.3.2. Precursors of 3 mercaptohexanol and 4 mercapto 4 methyl 2 grapes collected in the Spanish region of Catalonia in the Penedès pentanone Protected Designation of Origin (PDO) area. The vineyard (lat. 41° 25′ Four precursors of polyfunctional varietal mercaptans were ana 4.80″′ N; long. 01° 37′ 21.79″′ E) was non irrigated and goblet trained lyzed following the procedure validated by Concejero, Peña Gallego, Fernandez Zurbano, Hernández Orte, and Ferreira (2014): cysteine 4 solution of PFBBr in hexane, and letting the cartridge become imbibed mercapto 4 methyl 2 pentanone (CYSMP), glutathione 4 mercapto 4 with the reagent for 20 min at room temperature (25 °C). The remaining methyl 2 pentanone (GLUMP), cysteine 3 mercaptohexan 1 ol derivatizing agent was removed by addition of 100 µL of 2000 mg/L (CYSMH) and glutathione 3 mercaptohexan 1 ol (GLUMH). mercaptoglycerol in an aqueous solution containing 6.7% DBU, and letting the reaction take place for another 20 min at room temperature.

2.4. Chemical quantitative analysis of volatile compounds The cartridge was further rinsed with 4 mL of 0.2 M H3PO4 in water containing 40% methanol (v/v) and 1 mL of water. Derivatized analytes Several families of volatile compounds were analyzed in the wine were eluted with 600 μL of a solvent mixture (hexane 25% in diethyl with three different analytical methods. The analyses were performed ether), spiked with 10 μL of chromatographic internal standard (Octa in two different years but during the same period of each year to reduce fluoronaphthalene OFN 22.5 µL/L in hexane). The extract was finally potential variations associated with different post bottling times. 3 washed with 5 × 1 mL fractions of brine (200 g/L NaCl in water). 4 μL isobutyl 2 methoxypyrazine (IBMP) and 2 isopropyl 3 methoxypyr of this sample were directly injected in cold splitless mode into the GC azine (IPMP) were not analyzed since preliminary analysis revealed negative chemical ionization MS system. that these compounds were virtually absent from wines made with Carignan. 2.5. Statistical analysis

2.4.1. Major compounds (Liquid Liquid microextraction and GC FID Statistical analyses including regression tests were conducted with Analysis) Xlstat software (Addinsoft, Paris, France). The data were subjected to a The quantitative analysis of major compounds was carried out using three way analysis of variance (ANOVA) treatment (vintage × heating a validated and published method (Ortega, Lopez, Cacho, & Ferreira, temperature × heating time) with a first order interaction (n = 16; 2001). In accordance with this method, 3 mL of wine containing the residual degrees of freedom = 9). internal standards IS (2 butanol, 4 methyl 2 pentanol, 4 hydroxy 4 methyl 2 pentanone, and 2 octanol) and 7 mL of water were salted with 3. Results and discussion 4.5 g of ammonium sulfate and extracted with 0.2 mL of di chloromethane. The extract was then analyzed by GC with FID detec 3.1. Conventional enological parameters tion. The area of each analyte was normalized by that of its corre sponding IS and was then interpolated in the corresponding calibration Despite the pressing at high temperature that should have promoted plot built by applying exactly the same analytical method as that ap the extraction of amino acids from the pericarp (Geffroy et al., 2015; plied to synthetic wines containing known amounts of the analytes Poux, 1974), the content in yeast assimilable nitrogen (Table 1) re covering the natural range of occurrence of these compounds. Details flected by the sum of amino acids and ammonium was very low (less are given in the reference. than 100 mg/L for the prefermentive heat treatment at 50 °C for 30 min in 2014). To allow this content to surpass 150 mg/L and to avoid stuck 2.4.2. Minor compounds (SPE and GC Ion Trap MS Analysis) and sluggish fermentations, 300 mg/L of diammonium phosphate were This analysis was carried out using the method proposed by Lopez, added to all the flasks in 2014 and in 2015. Aznar, Cacho, & Ferreira (2002). In accordance with the method, 50 mL The three factors under study had a substantial effect on the con of wine, containing 25 μL of BHA solution and 75 μL of a surrogate ventional enological parameters as 9, 9 and 10 out of 13 measured standards solution (3 octanone, β damascone, heptanoic acid, and iso parameters were significantly impacted (at least at P < 0.05) by the propyl propanoate), were passed through a LiChrolut EN (Merck, vintage, the temperature and the heating time, respectively. Several Darmstadt, Germany) 200 mg cartridge at a rate of about 2 mL/min. significant interactions were noticed indicating that the extractability of The sorbent was dried under nitrogen stream (purity 99.999%). berry components under the effect of temperature and heating time is Analytes were recovered by elution with 1.3 mL of dichloromethane. complex and strongly dependent on the level of maturity of the grapes. 25 μL of an internal standard solution (4 hydroxy 4 methyl 2 penta The 2014 vintage was characterized by a dry spring, a mild summer none and 2 octanol, both at 300 mg per g of dichloromethane) were and rainy conditions during the ripening period. 2015 was an early added to the eluted sample. The extract was then analyzed by GC with vintage with warm dry weather conditions throughout the grapevine ion trap MS detection under the conditions described in the reference. vegetative cycle, especially in July when a severe 25 day heat wave occurred. Consequently the hot dry conditions in 2015 were more fa 2.4.3. Polyfunctional mercaptans (SPE and GC NCI MS Analysis) vorable to early maturity with higher sugar concentrations and lower This analysis was carried out using the method first proposed by levels of titratable acidity and malic acid. For this vintage, the extrac Mateo Vivaracho, Cacho, & Ferreira (2008) and further improved by tion rate was also lower which suggests that the berries were smaller Mateo Vivaracho, Zapata, Cacho, & Ferreira (2010). First, 0.2 g of and contained a lesser quantity of juice. Surprisingly, the phenolic ethylenediaminetetracetic acid and 0.6 g of L cysteine chlorohydrate concentrations in the experimental musts as reflected by anthocyanins were added to 25 mL of wine. This sample mixture was then transferred and TPI were weaker in 2015 than in 2014. It has been documented to a 20 mL volumetric flask where it was spiked with 15 μLofan that high temperature and increased water deficit up to a certain level ethanolic solution containing 1400 μg/L of 2 phenylethanethiol as in enhanced the phenolic accumulation in berries (Jackson & Lombard, ternal standard (IS). The complete volume was then transferred into a 1993). The excessive heat experienced in July 2015 after berry set with 24 mL screw capped vial together with 0.2 g of O methylhydrox temperatures largely above 30 °C might have contributed to lower an ylamine, shaken for 15 s, purged with pure nitrogen (99.999%), sealed thocyanin and proanthocyanidin synthesis (Mira de Orduña, 2010). In and incubated in a water bath at 55 °C for 45 min. Six milliliters of this addition, differences in the extractability of phenolic compounds at incubated sample were then loaded into a 50 mg Bond Elut ENV SPE harvest might have played a role. Distinct concentrations in must ni cartridge (Varian, Walnut Creek, USA). Major wine volatiles were re trogen and potassium contents might be the consequence of differences moved by rinsing with 4 mL of a 40% methanol water solution 0.2 M in in climatic conditions impacting the mineralization of soil organic phosphate buffer at pH 7.7. A second internal standard was also loaded matter and assimilation, or differences in fertilization practices between into the cartridge by passing it through 220 μL of solution (20 μLof4 the two vintages under study. methoxy α toluenethiol, 150 μg/L in ethanol and 200 μL water). Mer In most cases the heating temperature induced a significant increase captans retained in the cartridge were directly derivatized by passing in sugar concentration, pH, amino acids and ammonium. Differences in 1 mL of an aqueous solution of DBU (6.7%) and 50 μL of a 2000 mg/L the localization of organic acids, sugar and potassium have been Table 1 Results of three-way analysis of variance and impact of the heating conditions on skin to juice ratio and conventional enological parameters measured in musts.

Parameter 2014 2015 P-value

50 °C 75 °C 50 °C 75 °C Vintage (V) Temperature (Te) Time (Ti) V × Te V × Ti Te × Ti

30 min 3 h 30 min 3 h 30 min 3 h 30 min 3 h

Sugar concentration (°Brix) 21.5a 22.3 21.7 22.1 23.8 23.8 24.4 24.9 < 0.001 0.041 0.041 0.041 0.449 0.798 Titratable acidity (g/L tartaric 6.97 7.16 6.89 7.10 6.42 6.36 6.49 6.52 < 0.001 0.597 0.062 0.059 0.036 0.574 acid) pH 3.22 3.22 3.24 3.24 3.21 3.24 3.26 3.26 0.078 0.001 0.146 0.146 0.447 0.264 Tartaric acid (g/L) 3.91 3.01 2.84 3.54 4.64 3.64 2.94 2.81 0.073 < 0.001 0.005 < 0.001 0.027 < 0.001 Malic acid (g/L) 3.52 3.44 3.35 3.46 2.40 2.18 2.18 2.23 < 0.001 0.056 0.339 0.951 0.199 0.013 Potassium (g/L) 1.80 1.86 1.78 1.89 1.69 1.68 1.64 1.75 < 0.001 0.714 0.013 0.922 0.513 0.082 Amino acids (mg/L) 91 120 123 146 91 133 159 178 < 0.001 < 0.001 < 0.001 < 0.001 0.226 0.004 Ammonium (mg/L) 3.6 3.1 8.3 18.2 14.7 18.9 17.0 18.9 < 0.001 0.002 0.011 0.001 0.523 0.132 Anthocyanins (mg/L) 952 1448 1483 1329 495 1004 1232 1193 < 0.001 < 0.001 < 0.001 < 0.001 0.095 < 0.001 Total Phenolic Index (TPI) 51.7 68.0 70.4 80.8 33.2 55.2 67.5 80.8 < 0.001 < 0.001 < 0.001 < 0.001 0.059 0.005 Color Intensity 31.5 47.6 48.6 44.3 19.6 42.4 52.2 51.7 0.070 < 0.001 < 0.001 < 0.001 0.007 < 0.001 (A420 + A520 + A620) Extraction rate (%) 64.1 68.1 68.8 70.3 51.1 60.6 56.0 56.1 < 0.001 0.203 0.019 0.263 0.457 0.052

a Mean of two replicates. previously reported (Possner & Kliewer, 1985). Larger concentrations of tartaric acid has been precipitated by the excess of potassium. The fact glucose, fructose and potassium and lower levels of tartaric acid have that potassium was not lowered suggests that its extractability was been reported 120 days after anthesis in the skin and/or in the outer higher than those of tartaric under an increase of the heating time. part of the pulp in comparison with the inner part of the pulp and the It is important to emphasize that in 2014, the vintage with the area around the seeds. When crushing and pressing grapes, juices from higher extractability of phenolic compounds, the musts obtained from the center of the berries were first released. Then, through the fragili grapes heated at 50 °C for 3 h had a similar level of anthocyanins, TPI zation and destruction of cell membranes, the heating must have pro and color intensity as those treated at 75° for 30 min. This finding de moted the extraction of barely extractable juice from the periphery and monstrates that in some vintages the reduction of the heating tem the release of components from the skin into the musts. This led to an perature can be compensated for through an extension of the heating enrichment of the must in sugar and potassium and to a dilution in time. tartaric acid. The fact that no significant increase was observed for potassium can be explained by a higher precipitation of potassium 3.2. Aroma chemical composition tartrate which also contributed to intensify the decrease in tartaric acid and induced a clear increase in pH. In accordance with previous find Sixty four compounds from 11 chemical families were analyzed and ings (Geffroy et al., 2015; Poux, 1974), the heating temperature en detected in the 16 experimental wines produced. The results of the hanced the extraction of amino acids and to a lesser extent of ammo three way analysis of variance and the effect of the heating conditions nium contained in the pericarp. For anthocyanins, TPI and color on the aroma composition of the wines are shown in Table 2. intensity, the same increase was observed in most cases. Surprisingly, Of all the factors studied, the vintage had the greatest effect on the the musts obtained in 2014 from grapes heated at 50 °C for 3 h had a aroma chemical composition of the wines as 49 out of 64 parameters higher level of anthocyanins and color intensity than those treated at were impacted. It is commonly accepted that the vintage, owing to 75 °C for the same period, which suggests thermal degradation. Patras, varying climatic conditions, has a major influence on fruit quality and Brunton, O'Donnell, and Tiwari (2010) demonstrated that anthocyanins aroma composition. Without being exhaustive, several studies have were significantly affected in blackberry and strawberry puree by confirmed this observation analytically for white grape varieties thermal process treatments of 70 °C during holding times of 2 min. The (Schneider, Razungles, Augier, & Baumes, 2001) and for red wines reasons why this phenomenon was not observed in the 2015 must re made with traditional winemaking techniques (Kotseridis, Beloqui, main unclear. However, we can assume that the anthocyanin con Bertrand, & Doazan, 1998) and after a pre fermentative heat treatment centration in thermovinified musts is the result of extraction from the of the grapes (Geffroy et al., 2015). Under our experimental conditions, skin, thermal degradation and enzymatic degradations through poly the vintage factor significantly impacted most of the grape derived phenol oxidase. In our experimental conditions, we can assume that aroma compounds belonging to the terpenol and norisoprenoid, phenol, enzymatic oxidation usually inhibited from 50 °C to 60 °C was limited in vanillin derivate, cinnamate, polyfunctional mercaptan and lactone both temperature treatments. To suffer from degradation, anthocyanins chemical families. The reasons why changes in the fermentative aroma need first to be solubilized. The lesser extractability of anthocyanins in compounds (i.e. ethyl esters, acetates, acids and alcohols) were also 2015 as reflected by the levels in musts from grapes heated at 50 °C for noticed deserve further comment. Previous works highlighted that the 30 min might have contributed to reducing the time spent in solubilized amino acid composition in must had an impact on the production of form and therefore have limited their thermal degradation. fermentative volatile compounds in a model wine (Hernández Orte, As expected, the longer heating time improved the extraction rate Cacho, & Ferreira, 2002). It has been documented that the vintage and and generally enhanced the extraction of components located in the environmental conditions modified the amino acid profile of grapes and pericarp or in the outer part of the pulp such as sugars, potassium, that a high Yeast Assimilable Nitrogen (YAN) must promoted the pro amino acids, ammonium, TPI, anthocyanins and color intensity (except duction of ethyl esters, acetates, acids, carbonyl compounds and, in in 2014 at 75 °C for the reasons previously discussed). This could not be contrast, limited the synthesis of fusel alcohol (Bell & Henschke, 2005). always confirmed for tartaric acid, an acid involved together with po In accordance with these findings, the same conclusions could be drawn tassium in salt precipitation mechanisms and for which several first on our data in most cases in 2015, the vintage with the larger YAN order interactions were noticed. When the extension of the heating time level. led to a decrease in tartaric acid, we can suppose that the excess of 30 and 14 variables were significantly impacted for temperature Table 2 Results of three-way analysis of variance and impact of the heating conditions on the aroma composition of wines.

Parameter 2014 2015 P-value

50 °C 75 °C 50 °C 75 °C Vintage (V) Temperature (Te) Time (Ti) V × Te V × Ti Te × Ti

30 min 3 h 30 min 3 h 30 min 3 h 30 min 3 h

Ethyl esters (mg/L) Ethyl butyrate 0.05a 0.05 0.07 0.08 0.15 0.15 0.100 0.100 0.001 0.283 0.911 0.019 0.920 0.723 Ethyl 2-methylbutyrate 7.27 6.01 5.39 6.42 10.30 10.18 8.84 9.74 < 0.001 0.216 0.830 0.868 0.702 0.221 Ethyl isobutyrate 20.9 24.9 24.3 16.3 161.3 112.3 181.0 256.8 < 0.001 0.018 0.701 0.013 0.599 0.072 Ethyl hexanoate 0.19 0.23 0.20 0.20 0.25 0.21 0.59 0.62 0.001 0.002 0.834 0.002 0.773 0.865 Ethyl lactate 36.14 3.56 2.94 3.25 3.81 3.90 3.57 3.78 0.232 0.193 0.216 0.201 0.208 0.203 Ethyl octanoate 0.08 0.07 nd nd 0.24 0.16 0.22 0.26 < 0.001 0.627 0.675 0.114 0.834 0.324 Diethyl succinate 0.23 0.34 0.20 0.15 0.22 0.24 0.29 0.24 0.440 0.138 0.632 0.011 0.360 0.033 Ethyl isovalerate 6.3 75.1 84.0 47.0 4.54 21.7 87.7 69.4 0.367 < 0.001 0.343 0.027 0.310 0.001

Acetates (mg/L) Ethyl acetate 15.4 19.0 17.4 18.8 37.8 40.7 33.6 34.8 < 0.001 0.412 0.373 0.239 0.925 0.692 Hexyl acetate nd nd nd nd 0.022 0.021 0.007 0.007 < 0.001 0.013 0.916 0.013 0.916 0.898 Isoamyl acetate 0.37 0.22 0.37 0.54 1.01 0.91 0.41 0.51 0.025 0.238 0.920 0.026 0.915 0.362 Isobutyl acetate 9.7 7.1 10.9 15.4 12.9 11.1 6.1 7.4 0.429 0.904 0.833 0.016 0.728 0.166 Butyl acetate 2.0 2.7 2.0 2.8 14.5 14.9 13.7 13.9 < 0.001 0.542 0.443 0.511 0.733 0.978 Linalol acetate 0.64 0.63 0.69 0.88 1.74 1.81 1.68 1.68 < 0.001 0.742 0.456 0.152 0.729 0.680 Phenylethyl acetate 42.3 20.6 30.3 57.7 135.6 146.8 52.1 102.6 0.001 0.092 0.247 0.020 0.329 0.138

Acids (mg/L) Acetic acid 241 483 207 288 118 154 276 362 0.275 0.621 0.129 0.053 0.474 0.690 Butyric acid 0.27 0.24 0.26 0.32 0.43 0.44 0.32 0.36 0.001 0.215 0.372 0.018 0.913 0.170 Isobutyric acid 1.47 0.80 1.02 1.19 0.97 0.73 0.64 0.76 0.008 0.395 0.175 0.578 0.371 0.018 Isovaleric acid 1.67 1.43 1.12 1.61 1.12 0.86 0.73 0.87 < 0.001 0.077 0.751 0.968 0.364 0.014 Hexanoic acid 1.04 0.99 0.89 1.13 1.81 1.87 1.31 1.45 0.001 0.101 0.480 0.112 0.996 0.490 Octanoic acid 1.54 1.18 1.25 1.36 4.27 4.75 2.86 3.24 < 0.001 0.047 0.653 0.061 0.419 0.784 Decanoic acid 0.61 0.58 0.40 0.49 0.49 0.46 0.74 0.58 0.673 0.893 0.760 0.174 0.584 0.987

Alcohols (mg/L) Isoamyl alcohol 143 144 118 138 146 123 92 96 0.007 0.002 0.937 0.089 0.142 0.095 Benzyl alcohol 0.035 0.043 0.022 0.027 0.020 0.023 0.028 0.027 < 0.001 0.002 0.006 < 0.001 0.022 0.136 Methionol 1.89 0.92 0.59 0.91 1.99 1.57 0.65 1.00 0.035 < 0.001 0.073 0.132 0.138 < 0.001 1-hexanol 0.55 0.69 0.66 0.49 0.34 0.30 0.33 0.28 < 0.001 0.403 0.319 0.684 0.675 0.040 Cis-3-hexenol 0.045 0.054 0.047 0.041 0.065 0.052 0.058 0.050 0.001 0.038 0.047 0.743 0.018 0.280 1-butanol 0.48 0.65 0.53 0.58 0.73 0.80 0.76 0.76 < 0.001 0.753 < 0.001 0.752 0.032 0.008 β-phenylethanol 26.4 25.6 19.1 26.2 31.1 26.5 21.4 23.9 0.447 0.023 0.560 0.441 0.256 0.057 Isobutanol 30.1 27.2 25.6 30.8 12.5 10.4 8.8 9.9 < 0.001 0.263 0.740 0.456 0.467 0.026

Carbonyl compounds (mg/L) Diacetyl 2.3 3.5 2.5 1.7 nd 3.1 11.2 10.7 < 0.001 < 0.001 0.231 < 0.001 0.384 0.048 Syringaldehyde 0.99 0.97 0.73 0.62 0.39 nd 0.40 0.38 < 0.001 0.502 0.125 0.013 0.381 0.391 Acetoin 7.9 12.7 19.0 6.3 3.3 6.2 17.7 10.1 0.318 0.020 0.158 0.135 0.711 0.080 Acetaldehyde 9.9 12.9 9.7 6.9 4.3 6.8 22.6 16.6 0.002 < 0.001 0.233 < 0.001 0.198 < 0.001 Benzaldehyde 2.5 2.4 2.2 2.0 10.0 8.3 11.6 10.6 < 0.001 0.077 0.092 0.021 0.168 0.724

Terpenols and norisoprenoids (µg/L) Geraniol 56.3 51.7 44.7 50.2 5.6 8.4 5.8 5.3 < 0.001 0.022 0.619 0.115 0.820 0.270 β-citronellol 8.65 8.67 6.94 7.79 5.35 4.28 4.62 3.57 < 0.001 0.024 0.425 0.459 0.077 0.579 α-terpineol 6.89 6.01 6.05 8.42 2.84 3.02 3.18 3.95 < 0.001 0.025 0.047 0.783 0.617 0.006 Linalool 20.5 17.9 18.8 19.9 15.4 14.5 14.3 14.9 < 0.001 0.890 0.547 0.750 0.689 0.117 β-damascenone 5.42 3.21 2.43 1.11 6.53 5.21 3.97 2.34 < 0.001 < 0.001 < 0.001 0.594 0.379 0.380 α-ionone nd nd nd nd 0.027 0.060 0.049 0.078 0.004 0.493 0.290 0.493 0.290 0.945 β-ionone 0.32 0.29 0.32 0.39 0.33 0.34 0.35 0.31 0.990 0.364 0.878 0.226 0.494 0.574

Phenols (µg/L) Guaiacol 4.01 3.71 7.68 9.86 2.92 3.42 3.33 4.17 0.001 0.001 0.218 0.006 0.827 0.276 Eugenol 1.40 1.86 1.49 1.80 0.87 1.07 0.94 1.04 < 0.001 0.782 0.001 0.944 0.070 0.307 4-allyl-2,6-dimetoxyphenol 3.21 5.32 3.98 6.43 1.80 2.40 2.38 2.75 < 0.001 0.010 < 0.001 0.303 0.002 0.910 2,6-dimetoxyphenol 14.7 19.8 31.1 74.4 14.2 20.7 22.4 45.7 0.017 < 0.001 < 0.001 0.016 0.179 0.002 4-vinylguaiacol 17.6 25.4 25.0 41.5 17.1 23.3 27.6 37.1 0.470 < 0.001 < 0.001 0.909 0.163 0.069 4-vinylphenol 38.3 37.7 47.8 72.0 30.0 40.1 29.4 32.4 0.056 0.252 0.239 0.108 0.727 0.561 4-ethylphenol 0.055 0.079 0.180 0.190 0.149 0.194 0.193 0.211 0.044 0.020 0.370 0.128 0.787 0.699 o-cresol 0.32 0.41 0.17 0.37 0.30 0.26 0.31 0.28 0.507 0.418 0.253 0.271 0.091 0.509 m-cresol nd 1.46 1.44 1.53 nd nd nd nd < 0.001 0.012 0.011 0.012 0.011 0.020

Vanillin derivates (µg/L) Vanillin 5.06 2.64 2.79 2.61 6.20 3.85 4.33 4.87 0.001 0.026 0.005 0.250 0.519 0.002 Methyl vanillate 32.3 38.3 39.8 40.6 10.6 11.7 12.1 11.5 < 0.001 0.043 0.153 0.104 0.217 0.173 Ethyl vanillate 17.4 12.6 6.9 7.3 13.2 10.7 9.8 8.6 0.684 0.001 0.096 0.043 0.890 0.165 Acetovanillone 42.6 50.2 54.0 61.6 34.0 41.6 44.9 47.4 0.001 0.001 0.011 0.456 0.533 0.548

Cinnamates (µg/L) Ethyl dihydrocinnamate 0.28 0.29 0.39 0.52 0.80 0.69 0.54 0.57 0.001 0.865 0.756 0.008 0.361 0.265

Polyfunctional mercaptans (ng/L) (continued on next page) Table 2 (continued)

Parameter 2014 2015 P-value

50 °C 75 °C 50 °C 75 °C Vintage (V) Temperature (Te) Time (Ti) V × Te V × Ti Te × Ti

30 min 3 h 30 min 3 h 30 min 3 h 30 min 3 h

2-furfurylthiol 2.52 2.08 1.96 1.16 3.76 4.35 4.84 2.86 0.004 0.382 0.235 0.618 0.946 0.189 2-methyl-3-furanthiol 235 126 119 61 718 790 672 384 < 0.001 0.021 0.125 0.263 0.832 0.205 Benzylmercaptan 5.41 5.48 4.80 0.81 0.63 4.32 3.22 2.62 0.020 0.059 0.690 0.014 0.007 0.003 3-mercaptohexanol 2395 2188 1131 613 870 2220 2803 2045 0.106 0.259 0.886 0.001 0.176 0.025 Log(3-mercaptohexanol) 3.38 3.34 3.05 2.79 2.94 3.35 3.45 3.31 0.023 0.048 0.858 < 0.001 0.011 0.002 4-mercapto-4-methyl-2- 7.3 8.0 14.0 20.5 nd nd nd 11.2 < 0.001 0.001 0.016 0.230 0.530 0.023 pentanone 3-mercaptohexyl acetate 11.0 nd nd nd 11.0 35.0 19.7 15.3 0.006 0.291 0.674 1.000 0.154 0.396

Lactones (µg/L) γ-butirolactone 3.99 3.33 2.80 4.03 3.30 3.49 3.47 4.06 0.811 0.714 0.076 0.102 0.759 0.008 γ-nonalactone 2.15 4.75 5.62 6.05 1.02 1.37 2.23 3.12 < 0.001 0.001 0.020 0.265 0.274 0.314 nd. not detected. a Mean of two replicates. and heating time, respectively. This indicates that temperature has a as an enhancer of fruity aromas in red wine (Pineau et al., 2007). This greater effect than heating time on the aroma composition of the wines. aroma compound is produced from the neoxanthin carotenoid mostly In most cases, and especially for free aroma compounds extracted from found in the skin of mature grapes by both enzymatic and non enzy the berry skin or volatiles originating from precursors mainly located in matic reactions (Mendes Pinto, 2009). After an initial dioxygenase the pericarp, the extension of the heating period led to wines with a cleavage, the products undergo an enzymatic transformation to give the higher concentration by promoting their extraction. non aroma intermediate metabolites which are then converted through Surprisingly, the heating temperature did not induce any changes in acid catalyze into β damascenone. Despite a likely initial larger ex ethyl esters, acetates or acids with the exception of ethyl isobutyrate, traction of carotenoids from the pericarp under the heat effect, the ethyl hexanoate, ethyl isovalerate, hexyl acetate and octanoic acid. For denaturation of the carotenoid cleavage dioxygenase at 75 °C might the reasons previously discussed, the increase in YAN induced by this have contributed to limit the production of β damascenone from its factor should have promoted the production of most of these com precursor. The fact that this variable was also impacted by the heating pounds. This might be explained by the low initial nitrogen level of the time independently of the temperature tends to strengthen this suppo musts and the addition of diammonium phosphate (300 mg/L) to all the sition, and the hypothesis of thermal degradation can be discarded. It flasks, which contributed to reducing the percentage difference in YAN also suggests that the enzyme is sensitive to a temperature of 50 °C between the treatments under study. The impact of the temperature on when the heating is maintained for an extended period. carbonyls (diacetyl, syringaldehyde, acetaldehyde) remains difficult to The lower concentrations of 4 ethylphenol and m cresol combined discern as significant interactions involving the vintage and tempera with higher levels of guaiacol, 4 allyl 2,6 dimetoxyphenol, and 2,6 di ture factors were observed. For each of these compounds, an increase or metoxyphenol in wines from grapes heated at 75 °C reflects a thermal decrease recorded in the first year of study was followed by the opposite degradation of phenols. Indeed, the three latter compounds are known behavior in the second. end products of phenolic degradation (Michałowicz, & Duda, 2007). Differences in monoterpenols were noted between the temperature Vanillin, methyl vanillate, ethyl vanillate and acetovanillone are treatments. These aroma compounds originate from free and glycosi components originating from the ligneous material in the berries (Pearl, dically bound forms mainly located in the berry skin (Gomez, Martinez, 1942). As the grapes were destemmed, these compounds were mainly & Laencina, 1994). While the increase in temperature should have led extracted from the seeds. The increase in temperature should have to an overall increase in monoterpenols, the wines obtained from grapes promoted their extraction but this was not obvious in our experimental heated at 75 °C had a lower level of geraniol, β citronellol and a higher conditions. A significant decrease in vanillin and ethyl vanillate com concentration of α terpineol. Precursors (bound forms) are released bined with an increase in methyl vanillate and acetovanillone was ob through grape derived and yeast β glycosidase enzymes (Robinson served with an increase in temperature. et al., 2014). If the decrease in geraniol and β citronellol can be ex Despite major differences in 3 mercaptohexanol (3MH) concentra plained by the destruction of the endogenous enzyme activity by de tion between the treatments under study, none of the main factors had a naturation, the increase in α terpineol deserves further comments. significant impact on this key molecule involved in the varietal char Higher levels of this monoterpenol could be the consequence of the acter of wine. A significant interaction involving the temperature and thermal degradation of geraniol and β citronellol, a phenomenon pre vintage factors was observed and the residuals did not have a normal viously observed when grapes undergo a prefermentation heat treat distribution, which suggests the existence of multiplicative factors. As ment at 70 °C (Geffroy et al., 2015). As the odor thresholds of β ci proposed by Berry (1987), a logarithmic transformation of the variable tronellol (100 µg/L) and geraniol (30 µg/L) are lower than that of α was performed to convert the multiplicative session factor into an ad terpineol (250 µg/L), this degradation is likely to have an overall de ditive component. Data treatment using a 3 way ANOVA highlighted a preciative impact on the sensory characteristics of wine (Ferreira, significant effect of vintage and temperature with significant first order Lopez, and Aznar (2002)). interactions. In some experimental wines, 3MH concentrations ex The β damascenone content in red wine is generally around ceeded 2000 ng/L and were in the same range as those found in Sau 1 1.5 µg/L (Pineau, Barbe, Van Leeuwen, & Dubourdieu, 2007). The vignon blanc wines (Lund et al., 2009). Considering the perception concentrations found in our experimental wines, especially those made threshold of 60 ng/L for this aroma compound (Tominaga, Baltenweck from grapes heated at 50 °C, are to the best of our knowledge among the Guyot, Gachons, & Dubourdieu, 2000), 3MH should make an important highest ever reported in the literature. In accordance with previous contribution to wine aroma. In our experiment, the temperature had a research into thermovinification (Geffroy et al., 2015), the temperature clear depreciative impact on 3MH in 2014 while this could not be had a depreciative effect on β damascenone, which has been described confirmed in 2015. For a given grape material, the concentration of polyphenol oxidase (PPO) in the presence of oxygen which can favor the disappearance of varietal thiols (Roland et al., 2011). This non flavonoid phenolic compound is likely to be found at elevated levels in thermovinified wines and the inhibition of the tyrosinase (PPO) oc curring at 75 °C might also have impacted the 3MH concentrations. However, this phenomenon must have a small contribution in red wines as it was highlighted that their very high concentration of tannin should make these compounds very effective at scavenging quinone electro philes, and preventing the loss of thiols (Waterhouse, & Nikolantonaki, 2015). 4 Mercapto 4 methyl 2 pentanone (4MMP), another important contributor to varietal aroma whose perception threshold has been established at 0.8 ng/L (Tominaga et al., 2000), was systematically enhanced by a higher temperature and a longer heating time. As sug gested by Roland et al. (2011), this observation tends to indicate that enzymatic mechanisms leading to the production of 4MMP and 3MH differ. The difference in localization of 3MH and 4MMP precursors within the skin or pulp of the berry (Peyrot des Gachons, Tominaga, & Dubourdieu, 2002) might also play a role. Surprisingly, the increase in temperature led to a decrease in 2 methyl 3 furanthiol, an aroma compound imparting meaty notes formed through the thermal degradation of thiamin and through Maillard reactions (Bouchilloux, Darriet, & Dubourdieu, 1998). γ nonalactone, a scarcely studied aroma compound with an odor reminiscent of coconut (Nakamura, Crowell, Ough, & Totsuka, 1988), was also enhanced by a higher temperature and a longer heating time. As this odorant is produced from the yeast metabolism of amino and Fig. 1. Impact of the heating conditions on the precursors of 3-mercaptohexanol in 2014 keto acids (Etievant, 1991), this increase can be explained by the (□) and in 2015 (■). Error bars represent the standard error of the mean. greater extraction of amino acids from the pericarp. Concentrations analyzed in experimental wines are well under the sensory threshold of 3MH in wine is the result of the extraction of its precursors mostly from the molecule (30 µg/L) and the sensory contribution of this volatile the berry skin, their release by the yeast β lyase (Roland, Schneider, should remain weak. Razungles, & Cavelier, 2011) which depends on the fermentation con Before concluding, it is important to mention that our data are valid ditions (yeast strain, temperature, level of turbidity) and especially the for Carignan and might not be transposable to other grape cultivars. nitrogen composition of the must (Subileau, Schneider, Salmon, & Indeed, it has been shown that the extractability of skin compounds (i.e. Degryse, 2008). It was also recently proposed that 3MH precursors phenolic compounds, aroma precursors) from grapes that underwent a might be thermally degraded by a pre fermentation heat treatment of similar prefermentative heat treatment can differ between grape vari grapes (Geffroy et al., 2015). Levels of 3MH precursors in musts (Fig. 1) eties (Geffroy et al., 2015). However, our novel findings might stimu were much higher than those previously reported for Sauvignon blanc, late further studies on cultivars grown in wine regions where thermo a cultivar for which 3MH is an impact odorant (Capone, Sefton, vinification is popular. Another advantage for wineries of implementing Hayasaka, & Jeffery, 2010). This is likely to explain the substantial such heating would be to reduce the environmental impact and the cost levels of 3MH found in our experimental wines. As Cys3MH is mostly of the thermovinification process by saving energy both to heat the localized in the skin (Roland et al., 2011), we can suppose that the grapes and to cool down the musts after pressing. maceration at a temperature above 50 °C particularly enhanced the extraction of 3MH precursors in comparison with a typical white 4. Conclusion winemaking process which can include a short period of skin contact at a temperature below 20 °C. The analysis of 3MH precursors indicates The present study provides an in depth characterization of the effect that their extraction is significantly enhanced by the temperature at of temperature and heating time on the composition of two vintages of P < 0.001 for both compounds which allows us to reject the hypoth must and wine from Carignan. Among other findings, we have shown esis of thermal degradation. In our experiments carried out on the 2014 that, as expected, the heating temperature had a substantial impact on vintage, the extension of the heating period led to musts with a higher the extraction of amino acids and that probable thermal degradation of level of precursors while in 2015 the opposite change was observed. anthocyanins occurred when the grapes were heated at 75 °C. In the Both precursors were impacted by the vintage at P < 0.001 with case of the 2014 vintage, which had the higher extractability of phe higher concentrations found in 2015, the vintage with the higher ni nolic compounds, the reduction of the heating temperature was com trogen status, in accordance with previous findings (Choné et al. pensated for thanks to an extension of the heating time. The musts (2006)). No relationship could be identified between 3MH in wines and obtained from grapes heated at 50 °C for 3 h had a similar level of CYSMH (R2 = 0.01), GLUMH (R2 = 0.04), and the sum of CYSMH and phenolic compounds as those treated at 75° for 30 min. The heating GLUMH (R2 = 0.03) in musts. These discrepancies between the pre temperature had a greater effect than the heating time on the aroma cursors and the released thiols have been found previously by other composition of the wines. Surprisingly, the increase in temperature did authors (Peña Gallego, Hernández Orte, Cacho, & Ferreira, 2012). This not lead to wines with a higher level of volatile fermentative com indicates the eventual presence in must of other precursors and/or pounds. Several aroma compounds, especially grape derived molecules, suggests that differences in the aroma composition are likely to be ex were impacted by the rise in temperature, probably through degrada plained by distinct fermentation conditions such as the nitrogen or tion and enzyme denaturation mechanisms or as a consequence of the aminoacidic composition (Alegre, Culleré, Ferreira, & Hernández Orte, modification of fermentation conditions. Notably, wines made from 2017). Furthermore, trans caftaric acid is oxidized to o quinones by grapes treated at 50 °C had in most cases higher concentrations of geraniol, β citronellol, β damascenone and 3 mercaptohexanol. This indicates that the heating temperature can be modulated to produce Jackson, D. I., & Lombard, P. B. (1993). Environmental and management practices af- wines with a substantial concentration of grape derived volatiles. fecting grape composition and wine quality - A Review. American Journal of Enology and Viticulture, 44, 409–430. Kotseridis, Y., Beloqui, A. A., Bertrand, A., & Doazan, J. P. (1998). An analytical method Acknowledgements for studying the volatile compounds of Merlot noir clone wines. American Journal of Enology and Viticulture, 49,44–48. This study was carried out with financial support from the Lund, C. M., Thompson, M. K., Benkwitz, F., Wohler, M. W., Triggs, C. M., Gardner, R., ... Nicolau, L. (2009). New Zealand Sauvignon blanc distinct flavor characteristics: Communauté de Travail des Pyrénées (CTP) and received funding from Sensory, chemical, and consumer aspects. American Journal of Enology and Viticulture, the Région Midi Pyrénées and the Gobierno de Aragón. We are grateful 60,1–12. fi to Brigitte Mille, IFV, for her technical assistance and to the Celler Josep Marteau, G., & Olivieri, C. H. (1970). Bases et perspectives de la vini cation en rouge par macération à chaud. Bulletin Technique d’Information, 253. Piñol for the generous donation of the grapes. Mateo-Vivaracho, L., Cacho, J., & Ferreira, V. (2008). Improved solid-phase extraction procedure for the isolation and in-sorbent pentafluorobenzyl alkylation of poly- References functional mercaptans: Optimized procedure and analytical applications. Journal of Chromatography A, 1185,9–18. Mateo-Vivaracho, L., Zapata, J., Cacho, J., & Ferreira, V. (2010). Analysis, occurrence, Alegre, Y., Culleré, L., Ferreira, V., & Hernández-Orte, P. (2017). Study of the influence of and potential sensory significance of five polyfunctional mercaptans in white wines. varietal amino acid profiles on the polyfunctional mercaptans released from their Journal of Agricultural and Food Chemistry, 58, 10184–10194. precursors. Food Research International, 100, 740–747. Mendes-Pinto, M. M. (2009). Carotenoid breakdown products the norisoprenoids in wine Auw, J., Blanco, V., O'Keefe, S., & Sims, C. (1996). Effect of processing on the phenolics aroma. Archives of Biochemistry and Biophysics, 483, 236–245. and color of Cabernet Sauvignon, Chambourcin, and Noble wines and juices. Michałowicz, J., & Duda, W. (2007). Phenols transformations in the environment and American Journal of Enology and Viticulture, 47, 279–286. living organisms. Current Topics in Biophysics, 30,24–36. Bell, S.-J., & Henschke, P. A. (2005). Implications of nitrogen nutrition for grapes, fer- Mira de Orduña, R. (2010). Climate change associated effects on grape and wine quality mentation and wine. Australian Journal of Grape and Wine Research, 11, 242–295. and production. Food Research International, 43, 1844–1855. Berry, D. A. (1987). Logarithmic Transformations in ANOVA. Biometrics, 43, 439–456. Moreno, J., Medina, M., & Garcia, M. (1988). Optimization of the fermentation conditions Bouchilloux, P., Darriet, P., & Dubourdieu, D. (1998). Identification d’un thiol fortement of musts from Pedro Ximénez grapes grown in Southern Spain. Production of higher odorant, le 2-méthyl-3-furanthiol, dans les vins. Vitis, 37, 177–180. alcohols and esters. South African Journal of Enology and Viticulture, 9,16–20. Capone, D. L., Sefton, M. A., Hayasaka, Y., & Jeffery, D. W. (2010). Analysis of precursors Nakamura, S., Crowell, E., Ough, C., & Totsuka, A. (1988). Quantitative analysis of γ- to wine odorant 3-mercaptohexan-1-ol using HPLC-MS/MS: Resolution and quanti- nonalactone in wines and its threshold determination. Journal of Food Science, 53, tation of diastereomers of 3-S-cysteinylhexan-1-ol and 3-S-glutathionylhexan-1-ol. 1243–1244. Journal of Agricultural and Food Chemistry, 58, 1390–1395. Patras, A., Brunton, N. P., O'Donnell, C., & Tiwari, B. K. (2010). Effect of thermal pro- Choné, X., Lavigne-Cruège, V., Tominaga, T., van Leeuwen, C., Castagnède, C., Saucier, cessing on anthocyanin stability in foods; mechanisms and kinetics of degradation. C., & Dubourdieu, D. (2006). Effect of vine nitrogen status on grape aromatic po- Trends in Food Science & Technology, 21,3–11. tential: Flavor precursors (S-cysteine conjugates), glutathione and phenolic content in Pearl, I. A. (1942). Vanillin from lignin materials. Journal of the American Chemical Society, Vitis vinifera L. Cv Sauvignon blanc grape juice. Journal International des Sciences de la 64, 1429–1431. Vigne et du Vin, 40,1–6. Peña-Gallego, A., Hernández-Orte, P., Cacho, J., & Ferreira, V. (2012). S-Cysteinylated Concejero, B., Peña-Gallego, A., Fernandez-Zurbano, P., Hernández-Orte, P., & Ferreira, and S-glutathionylated thiol precursors in grapes. A review. Food Chemistry, 131, V. (2014). Direct accurate analysis of cysteinylated and glutathionylated precursors 1–13. of 4-mercapto-4-methyl-2-pentanone and 3-mercaptohexan-1-ol in must by ultrahigh Peyrot des Gachons, C., Tominaga, T., & Dubourdieu, D. (2002). Localization of S-cysteine performance liquid chromatography coupled to mass spectrometry. Analytica Chimica conjugates in the berry: Effect of skin contact on aromatic potential of Vitis vinifera L. Acta, 812, 250–257. cv. Sauvignon blanc must. American Journal of Enology and Viticulture, 53, 144–146. Cottereau, P., & Desseigne, J. M. (2007). Chauffage de la vendange et arômes fruités. In Pineau, B., Barbe, J.-C., Van Leeuwen, C., & Dubourdieu, D. (2007). Which impact for β- Proceedings of the technical seminar Entretiens vitivinicoles Rhône Mediterranée, damascenone on red wines aroma? Journal of Agricultural and Food Chemistry, 55, Narbonne, France (IFV: Le Grau-du-Roi, France) (pp. 20–22). 4103–4108. Doco, T., Williams, P., & Cheynier, V. (2007). Effect of flash release and pectinolytic Possner, D. R. E., & Kliewer, W. M. (1985). The localisation of acids, sugars, potassium enzyme treatments on wine polysaccharide composition. Journal of Agricultural and and calcium in developing grape berries. Vitis, 24, 229–240. Food Chemistry, 55, 6643–6649. Poux, C. (1974). Chauffage de la vendange et composés azotés. Industries Alimentaires et Etievant, P. (1991). Wine. Volatile compounds in foods and beverages (pp. 483–546). . Agricoles, 91, 335–340. Ferreira, V., Lopez, R., & Aznar, M. (2002). Olfactometry and aroma extract dilution Ribéreau-Gayon, P. (1970). Les dosages des composés phénoliques totaux dans le vin analysis of wines. Analysis of taste and aroma (pp. 89–122). Heidelberg: Springer, rouge. Chimie Analitica, 52, 627–631. Berlin. Ribéreau-Gayon, P., & Stonestreet, E. (1965). Le dosage des anthocyanes dans le vin Fischer, U., Strasser, M., & Gutzler, K. (2000). Impact of fermentation technology on the rouge. Bulletin de la Société de Chimique de France, 9, 2649–2652. phenolic and volatile composition of German red wines. International Journal of Food Ribéreau-Gayon, P., Dubourdieu, D., Donèche, B., & Lonvaud, A. (2005). Handbook of Science & Technology, 35,81–94. enology, the microbiology of wine and vinifications. John Wiley & Sons. Geffroy, O., Lopez, R., Serrano, E., Davaux, F., Gracia-Moreno, E., Cacho, J., & Ferreira, V. Robinson, A. L., Boss, P. K., Solomon, P. S., Trengove, R. D., Heymann, H., & Ebeler, S. E. (2014). Macération préfermentaire à chaud : Modulation du profil sensoriel des vins (2014). Origins of grape and wine aroma. Part 1. Chemical components and viti- rouges par le niveau de turbidité des moûts et la température de fermentation. Revue cultural impacts. American Journal of Enology and Viticulture, 65,1–24. des Œnologues, 150,18–20. Roland, A., Schneider, R., Razungles, A., & Cavelier, F. (2011). Varietal thiols in wine: Geffroy, O., Lopez, R., Serrano, E., Dufourcq, T., Gracia-Moreno, E., Cacho, J., & Ferreira, Discovery, analysis and applications. Chemical Reviews, 111, 7355–7376. V. (2015). Changes in analytical and volatile compositions of red wines induced by Roujou de Boubée, D. (2000). Recherche sur la 2-méthoxy-3-isobutylpyrazine dans les raisins pre-fermentation heat treatment of grapes. Food Chemistry, 187, 243–253. et les vins. Approches analytique, biologique et agronomique. Ph.D. thesis. Bordeaux, Geffroy, O., Siebert, T., Silvano, A., & Herderich, M. (2016). Impact of winemaking France: University of Bordeaux 2 170 pp. techniques on classical enological parameters and rotundone in red wine at the la- Schneider, R., Razungles, A., Augier, C., & Baumes, R. (2001). Monoterpenic and nor- boratory scale. American Journal of Enology and Viticulture, 6, 141–146. isoprenoidic glycoconjugates of Vitis vinifera L. cv. Melon B. as precursors of odorants Girard, B., Kopp, T. G., Reynolds, A. G., & Cliff, M. (1997). Influence of vinification in Muscadet wines. Journal of Chromatography A, 936, 145–157. treatments on aroma constituents and sensory descriptors of Pinot noir wines. Subileau, M., Schneider, R., Salmon, J. M., & Degryse, E. (2008). Nitrogen catabolite American Journal of Enology and Viticulture, 48, 198–206. repression modulates the production of aromatic thiols characteristic of Sauvignon Gomez, E., Martinez, A., & Laencina, J. (1994). Localization of free and bound aromatic Blanc at the level of precursor transport. FEMS Yeast Research, 8, 771–780. compounds among skin, juice and pulp fractions of some grape varieties. Vitis, Tominaga, T., Baltenweck-Guyot, R., Gachons, C. P. D., & Dubourdieu, D. (2000). 33,1–4. Contribution of volatile thiols to the aromas of white wines made from several Vitis Hernández-Orte, P., Cacho, J. F., & Ferreira, V. (2002). Relationship between varietal vinifera grape varieties. American Journal of Enology and Viticulture, 51, 178–181. amino acid profile of grapes and wine aromatic composition. Experiments with model Waterhouse, A. L., & Nikolantonaki, M. (2015). Quinone reactions in wine oxidation. solutions and chemometric Study. Journal of Agricultural and Food Chemistry, 50, Advances in Wine Research (pp. 291–301). American Chemical Society. 2891–2899.